CN101329763B - Method for segmenting structures in image data records and image processing unit - Google Patents
Method for segmenting structures in image data records and image processing unit Download PDFInfo
- Publication number
- CN101329763B CN101329763B CN2008101253079A CN200810125307A CN101329763B CN 101329763 B CN101329763 B CN 101329763B CN 2008101253079 A CN2008101253079 A CN 2008101253079A CN 200810125307 A CN200810125307 A CN 200810125307A CN 101329763 B CN101329763 B CN 101329763B
- Authority
- CN
- China
- Prior art keywords
- image data
- segmentation result
- zone
- mark
- apart
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/174—Segmentation; Edge detection involving the use of two or more images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/24—Indexing scheme for image data processing or generation, in general involving graphical user interfaces [GUIs]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20092—Interactive image processing based on input by user
- G06T2207/20096—Interactive definition of curve of interest
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20092—Interactive image processing based on input by user
- G06T2207/20104—Interactive definition of region of interest [ROI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
- Processing Or Creating Images (AREA)
Abstract
The present invention relates to a method for dividing structure in an image data set and an image processing unit for implementing the method. In the method, a first segmentation of a structure in an image data set is firstly carried out, and a first final segmentation result is obtained therefrom. A region in the image data set is selected based on the first final segmentation result obtained. A first band is placed at a first, outwardly pointing distance (d1) from the selected region. This first band characterizes a background region. A second band is placed at a second, inwardly pointing distance (d2) from the projected first final segmentation result of the first segmentation. This second band characterizes a structure region. A further segmentation is carried out based on the characterized background region and the characterized structure region, and the final segmentation result of the further segmentation is saved and/or displayed.; Furthermore, an image processing unit for carrying out the method is disclosed.
Description
Technical field
The present invention relates to a kind of for method and a kind of graphics processing unit for carrying out the method at the image data set segmenting structure.
Background technology
To being a part and parcel for the analysis of view data (particularly medical image) cutting apart of anatomical structure.Important example has: to the planning of surgical intervention, and the inspection of the cubing of organ, to the judgement of metabolic process, the perhaps statistical modeling of organ.
At this, about the image information of three-dimensional (3D) structure, normally occur according to the sequence such as two dimension (2D) scanning slice of the image mode of computed tomograph (CT) or magnetic resonance tomography instrument (MRT).Therefore, usually must at first be partitioned into the desirable structure in each scanning slice, can comprehensively be the 3D structure with it just then.But, also there has been the partitioning algorithm of directly cutting apart for to the 3D structure.
Almost can not realize cutting apart for the full automatic of medical image according to present technical merit.Therefore unavoidable is that the user must be able to intervene independently during cutting apart.According to the difference of the partitioning algorithm that adopts, the user was cut apart according to the different mode interventions, in order to control the result according to desirable mode.
There is the different algorithms that is used for interactive segmentation.At this, can distinguish inter alia two different main group: based on the partitioning algorithm of profile and based on the partitioning algorithm in zone.As the exemplary partitioning algorithm based on profile, at this so-called Livewire method is described momently.The method has been established as from 3D data group organ structure high-quality interactive segmentation qualitatively.It particularly is extremely reliable under the condition of strong picture contrast.Yet, under the condition of weak picture contrast, often need the too much intervention of user in order to realize the acceptable result.
The key concept of Livewire method is as follows: starting point of (for example by means of cursor or mouse) mark on the profile of the structure of user in the image data set that illustrates, then with cursor movement to another position in this image data set.The Livewire algorithm calculates the variation of the profile of the position from starting point to current cursor.Adopt so-called cost function, this function allows to calculate the path of satisfying best specified criteria for this reason, and is for example minimum along the change of the gradient absolute value in this path.If the path of this calculating correctly is not positioned on the profile of structure, then the user can be for example by simple click and mobile intervention to this path with proofreading and correct.More specifically, for example at " the Interactive Live-Wire Boundary Extraction " of W.A.Barret and E.N.Mortensen (1997), Medical Image Analysis, Vol.1 is described among No.4.pp.33 1341.CVPR ' the 99#107 page 6/7.
As for the example based on the partitioning algorithm in zone, so-called GraphCut method is described momently.The method also has been established as high-quality qualitatively, and has also realized good result under the condition of weak picture contrast.The user marks the image-region that is positioned at inside configuration and the image-region that is positioned at structural outer in the GraphCut method.Also by means of the uncontinuity of the maximum between these zones of cost function calculation, this cost function comprises for example gray value information in the zone of institute's mark as criterion to the GraphCut method.This maximum uncontinuity is corresponding to the border of structure.If the result after cutting apart first is also unsatisfactory, then can mark other inside and outside zone, until present acceptable segmentation result.For example in US 2004/0008886A1, provided the more specifically description for GraphCut algorithm and relevant cost function.
By means of this partitioning algorithm, the user operates by given 3D rendering data group layer by layer, until be partitioned into total.According to the partitioning algorithm that adopts and the difference of the picture contrast that occurs separately, the user may be forced to intervene frequently, and this point has prolonged the processing time to 3D rendering data group significantly.
Therefore, also have the demand for user-friendly partitioning algorithm, this partitioning algorithm allows cutting apart rapidly structure as far as possible intuitively and along with few reciprocation.
Summary of the invention
Therefore, the technical problem to be solved in the present invention is, a kind of method be used to cutting apart is provided, and on few intuitively operation steps, these operation stepss are so that can effectively operate with the interaction constraints of needed user and dividing method for the method.
Above-mentioned technical matters is to solve for the method at the image data set segmenting structure by a kind of, and described method has following steps: carry out in image data set first of structure is cut apart; First obtain the first segmentation result cutting apart from this; In this image data set, selecting a zone on the basis of this first segmentation result; By the first band being set and the mark background area apart from selected zone the first distance toward the outer side; Mark structure is regional by at the second distance of the selected zone of distance towards the inboard the second band being set; In the background area of institute's mark and the basis of the structural region of institute's mark carry out and further cut apart; Show and/or store the segmentation result that this is further cut apart.
This according to method of the present invention in, at first in image data set, carry out first of structure cut apart and therefrom obtained the first segmentation result.This segmentation result can or be in the 2D layer etc. profile (Isokontur), i.e. a heavy diversity (1-Mannigfaltigkeit), or be in 3D data group homalographic (
), be dual diversity (2-Mannigfaltigkeit).The below will be independent of dimension ground and describe the method.
On the basis of the first segmentation result that obtains, in this image data set, select a zone.With the first band be arranged on the selected zone of distance toward the outer side first apart from the d1.This first tape label the background area.
At this, different according to dimension, band is interpreted as respectively parallel profile (expression one heavy diversity) or the homalographic (representing dual diversity) of waiting.
The second band is arranged on the selected zone of distance on the second distance d2 of inboard.This second tape label structural region.In the background area of institute's mark and the basis of the structural region of institute's mark carry out and further cut apart, and show and/or store the segmentation result that this is further cut apart.That is to say " inside " or " outside " of the structure that concept " inboard " or " outside " expression are to be split.
The setting of band is automatically realized on the basis of the bandwidth of distance d1 and d2 and the first band and the second band by segmentation procedure, corresponding to the situation of image data set these distance and bandwidth given in advance.Usually, different according to resolution and image situation, the width of band between 1 and 20 pixels or more between.Distance is selected according to situation.Usually, distance is similarly several pixels, 3 to 5 pixels for example.
So that greatly simplified the cutting apart of structure, and reduce needed user's interactive quantity according to method of the present invention, because automatically carry out initialization to further cutting apart.This is further cut apart to have finished and cuts apart.
Can will be used in the different situations that the 3D rendering data component is cut according to method of the present invention advantageously.
On the one hand, has advantage in successively the cutting apart of 3D rendering data group.At this, with the first segmentation result of cutting apart in the ground floor of 3D rendering data group, project on the second layer adjacent with this ground floor of this 3D rendering data group.This Projection selection a zone in this second layer, in order to or band is placed in one.
By in the second layer adjacent with this ground floor to the automatic mark of background area and structural region, can in this second layer, for example utilize the GraphCut method promptly and under the condition that does not almost have further customer interaction, carry out and cut apart.Therefore, can carry out in the short period of time cutting apart successively.In the following description this way is called " band method (Ribbon Method) ".
On the other hand, can improve in a simple manner the surface of 3D rendering structure comprehensive from a plurality of layers segmentation result.For this reason, as the first segmentation result, will be from the segmentation result of successively cutting apart the surface of comprehensive 3D structure, be considered as dual multifarious (zweimannigfaltig) surface and be chosen as such zone, in order to or according to mode described above the first band and second is with and is placed in one.
Therefore cut apart (for example GraphCut method) by the method is initialized, " inboard " and the maximum uncontinuity between " outside " in the 3D structure are provided, and provide a kind of and be corrected and level and smooth surface.In the following description this way is called " 3D band method (Ribbon Method3D) ".
In addition, technical matters of the present invention still solves by a kind of graphics processing unit, and this graphics processing unit comprises: input block is used for input command; Display unit is used for the display image data group; Storage unit is used for storing and load image data group; And calculation element, utilize this graphics processing unit can implement according to method of the present invention.
Description of drawings
Other advantage of the present invention and details provide by the embodiment that the following describes and by means of accompanying drawing.The embodiment that provides does not limit the present invention.In the accompanying drawing:
Fig. 1 shows the indicative flowchart according to method of the present invention,
Fig. 2 shows the possibility of cutting apart with the example of the strong simplification of structure to be split in the 2D image data set,
Fig. 3 shows the explanation according to method of the present invention with the example among Fig. 2,
Fig. 4 shows the schematic diagram by the section of 3D structure comprehensive after the cutting apart fully of relevant to all layer,
Fig. 5 shows divided 3D structure after aftertreatment,
Fig. 6 shows for the schematic diagram of carrying out according to the graphics processing unit of method of the present invention.
Embodiment
Fig. 1 shows explanation according to the indicative flowchart of the flow process of method of the present invention.At this, in image data set, carry out first and cut apart (square frame 11), in order to obtain the first segmentation result (square frame 13).
On the basis of resulting segmentation result, in this image data set, select a zone (square frame 15).
Around selected region division the first band and the second band, so that mark background area and structural region (square frame 17).(square frame 19) further cut apart on basis in the zone of institute's mark.At last, show and/or store the segmentation result (square frame 21) that this step is cut apart.Below at first be explained in more detail this method about Fig. 3 and Fig. 4.
Fig. 2 has represented the data group 31 of strong simplification, and it comprises structure to be split 33 and another structure 35.
Point A and B have marked the point that structure 33, another structure 35 and background are had a common boundary.
For segmenting structure 33, unique partitioning algorithm that user or can select is cut apart for whole perhaps can have so preferred possibility: utilize different ground, partitioning algorithm region-by-region to process image data set, as below will as described in.
If user selection a kind of partitioning algorithm based on the zone, in order to for example in following zone, structure 33 is cut apart, namely, this structure and another structure 35 are had a common boundary in this zone, what for example go out as shown is labeled as rectangular area 32 (dotted line) like that, and extend on the border in this rectangular area between the structure 33 and 35.This point can be by means of such as the input media that comprises mouse, carries out with the selection as the rectangular selection instrument of cursor according to common mode.
In addition, for the initialization based on the partitioning algorithm (for example GraphCut method) in zone, the user also selects zone 34 as the structural region 34 that belongs to structure 33 in the inside of structure to be split 33, and select as a setting zone 36, zone 36 in the outside of structure 33, wherein, in this association, all are not belonged to the background that is labeled as of structure 33 to be split.This point realizes by means of input media again, wherein also can provide other known selection tool as cursor for.
At this moment, the partitioning algorithm based on the zone after the initialization splits structure 33 in the selected zone 32 that surrounds by white dashed line.If the segmentation result in the zone 32 of institute's mark also is not no problem, then the user can be structural region or background area with other zone marker, until cutting apart of hope occurred meeting in zone 32.As this segmentation result of cutting apart, black solid line between the zone 33 and 35, extend to a B from an A according to clockwise direction has been shown in this example.
In order further to cut apart this structure to be split 33, a kind of partitioning algorithm based on profile of user selection this moment, for example Livewire method.For this zone to be split of mark, the user clicks a starting point at the profile of structure to be split 33 simply, clicking point B for example, and utilize cursor to be moved further along structure 33, be roughly according to clockwise direction at this.
At this, always calculate and show the segmentation result of the position from starting point B to cursor.This utilization based on the partitioning algorithm of profile cut apart during, the user can be on desired profile other point of mark as the point of fixity that is used for partitioning algorithm.
Thus, the user moves forward until arrive the terminal point of for example putting A.On this terminal point, the user finishes the mark for zone to be split, and side by side finishes to utilize based on the cutting apart of the partitioning algorithm of profile, for example by double-click on or with another mouse button click.Show the segmentation result that this is cut apart by the solid line that extends to the black of an A along clockwise direction from a B in this example.
If the user is also dissatisfied for relevant segmentation result, then it can reselect a kind of partitioning algorithm, and structure 33 does not also make its satisfactorily divided zone with proofreading and correct in order to again cut apart wherein.At this, the user also can adopt manual cutting apart.In addition, the user also can utilize another kind of partitioning algorithm that divided zone is cut apart again at this where necessary.In addition, also can adopt known corrective action, for example so-called " the Path Cooling " in the Livewire method.
If the user splits structure 33 completely for the segmentation result satisfaction that realizes and its, then it can be stored the segmentation result of realizing as the segmentation result of shown layer.
Fig. 3 in Fig. 2 with the example images data group 31 of structure to be split 33 and another structure 35 as example, illustrated according to method of the present invention, i.e. so-called " band method ".
The segmentation result 38 of adjacent layer is projected in the image data set 31, and select thus a zone.With first with 37 be arranged on apart from the segmentation result 38 of the adjacent layer of institute's projection toward the outer side first apart from the d1.At this, this first tape label background area 36.
Be arranged on segmentation result 38 apart from the adjacent layer of institute's projection with second on the second distance d2 of inboard with 39.At this, this first tape label structural region 34.
Suitably given in advance apart from d1 and d2 and first with 37 and second with 39 width.
After this automatic initialization based on the partitioning algorithm in zone, the user can continue cutting apart structure 33 as described above.
Utilize this mode greatly to simplify in 3D rendering data group cutting apart successively to structure, because omitted manual initialization.Thus, the user can extremely promptly pass through a plurality of layers of ground operation of the 3D rendering data group of multilayer.
If structure clash is a plurality of parts in certain layer, then respectively single part is cut apart, and subsequently they for example are comprehensive element by means of how much (Constructive solid geometry, CSG) unifications of constructive entity.This method for example is known as " division (Split) " and " merging (Merge) " method.
Among Fig. 4 by passing through one by the section of the comprehensive 3D structure 53 of six single divided layers with the area of shade is schematically illustrated.Limiting the solid line 58 of the black of this comprehensive 3D structure 53, is to be undertaken comprehensively forming by the segmentation result with the edge of single layer and layer.Therefore, line 58 expression is by the section on this comprehensive surface and the selected zone that is illustrated in thus this structure 53.
In addition, for " 3D band method " is described, shows in the outside of relevant section what toward the outer side distance D 1 arranged and first be with 57 (its mark background area).Similarly, this section also be included on the distance D 2 of inboard arrange second be with 59.
Again suitably distance D 1 given in advance and D2 and first with 57 and second with 59 width.
By means of based on the partitioning algorithm in zone, for example utilize the GraphCut method, can on the basis in the zone of institute's mark, again cut apart the dual multifarious surface of comprehensive structure 53.Thisly cut apart the uncontinuity that the maximum between " inboard " and " outside " in 3D rendering data group is provided, and the surface after the correction of 3D structure to be split is provided thus.
Fig. 5 schematically show cut apart and other by abundant known aftertreatment (for example, level and smooth and fidelity) section by the 3D structure among Fig. 4 afterwards.Obtained level and smooth surface 68.
At last, Fig. 6 shows for the schematic diagram of carrying out according to the graphics processing unit 40 of method of the present invention.
Graphics processing unit 40 comprises: input media 41 is used for for example by means of mouse or keyboard input command; Display device 42 is used for the display image data group; Memory cell 43 is used for storage and load image data group; And calculation element 44, be used for carrying out and calculate.
Realized that at graphics processing unit 40 component cuts algorithm, these partitioning algorithms are reciprocally combination on function corresponding to the present invention.
Claims (8)
1. method that is used at the image data set segmenting structure, described method has following steps:
-in image data set, carry out first of structure is cut apart,
-first obtain the first segmentation result cutting apart from this,
-in this image data set, selecting a zone on the basis of this first segmentation result,
-pass through the first band is being set and the mark background area apart from selected zone the first distance (d1, D1) toward the outer side,
-mark structure is regional by at the second distance (d2, D2) of the selected zone of distance towards the inboard the second band being set,
-in the background area of institute's mark and the basis of the structural region of institute's mark carry out and further cut apart,
-show and/or store the segmentation result that this is further cut apart.
2. method according to claim 1 wherein, is utilized based on the partitioning algorithm execution in zone is described and is further cut apart.
3. method according to claim 1 and 2, wherein, described the first band and/or the second band have the width that can be scheduled to.
4. method according to claim 1 and 2 wherein, is carried out described first and is cut apart in the ground floor of three-dimensional image data sets, and is chosen in selected zone on the basis of this first segmentation result in the second layer adjacent with this ground floor.
5. method according to claim 4, wherein, this selected zone is the projection of described the first segmentation result on the described second layer.
6. method according to claim 5, wherein, described first cut apart be in three-dimensional image data sets to successively the cutting apart of structure, and described the first segmentation result is the surface of this structure that combination obtains from the segmentation result of successively cutting apart.
7. method according to claim 6, wherein, this selected zone is the body structure surface that combination obtains in described three-dimensional image data sets.
8. one kind is used in the image measurement unit of image data set segmenting structure, has:
-be used at first device cut apart of image data set execution to structure,
-be used for first cutting apart the device that obtains the first segmentation result from this,
-for the device of on the basis of this first segmentation result, selecting a zone in this image data set,
-be used for passing through the first band being set and the device of mark background area apart from selected zone the first distance (d1, D1) toward the outer side,
-be used for being with and the device in mark structure zone by arranging second at the second distance (d2, D2) of the selected zone of distance towards the inboard,
-be used in the background area of institute's mark and the device of further cutting apart is carried out on the basis of the structural region of institute's mark,
-for the device that shows and/or store this segmentation result of further cutting apart.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| DE102007028895.8 | 2007-06-22 | ||
| DE102007028895A DE102007028895B4 (en) | 2007-06-22 | 2007-06-22 | Method for segmenting structures in 3D image data sets |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101329763A CN101329763A (en) | 2008-12-24 |
| CN101329763B true CN101329763B (en) | 2013-01-02 |
Family
ID=40075834
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008101253079A Expired - Fee Related CN101329763B (en) | 2007-06-22 | 2008-06-18 | Method for segmenting structures in image data records and image processing unit |
Country Status (3)
| Country | Link |
|---|---|
| US (1) | US8180151B2 (en) |
| CN (1) | CN101329763B (en) |
| DE (1) | DE102007028895B4 (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8842909B2 (en) * | 2011-06-30 | 2014-09-23 | Qualcomm Incorporated | Efficient blending methods for AR applications |
| CN104781649B (en) * | 2012-11-16 | 2019-07-05 | 贝克曼考尔特公司 | The evaluation system and method for flow cytometry data segmentation result |
| DE102015206701A1 (en) * | 2015-04-15 | 2016-10-20 | Siemens Healthcare Gmbh | Method and user interface for segmenting an object and imaging device, computer program product and machine-readable data carrier |
| CN108765428A (en) * | 2017-10-25 | 2018-11-06 | 江苏大学 | A kind of target object extracting method based on click interaction |
| CN110009684A (en) * | 2019-04-15 | 2019-07-12 | 云南民族大学 | One kind is for sleeper localization method under tamping operation |
| CN110110617B (en) * | 2019-04-22 | 2021-04-20 | 腾讯科技（深圳）有限公司 | Medical image segmentation method and device, electronic equipment and storage medium |
| CN113129307A (en) * | 2019-12-31 | 2021-07-16 | 华为技术有限公司 | Image segmentation method and image segmentation device |
| CN111768393A (en) * | 2020-07-01 | 2020-10-13 | 上海商汤智能科技有限公司 | Image processing method and device, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1524247A (en) * | 2001-05-17 | 2004-08-25 | О | A variational approach for the segmentation of the left ventricle in mr cardiac images |
| CN1745715A (en) * | 2004-09-09 | 2006-03-15 | 西门子公司 | Method for segmenting anatomical structures from 4d image data |
| CN1748641A (en) * | 2004-08-09 | 2006-03-22 | 西门子公司 | Method for segmenting a medical data set |
Family Cites Families (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6078688A (en) | 1996-08-23 | 2000-06-20 | Nec Research Institute, Inc. | Method for image segmentation by minimizing the ratio between the exterior boundary cost and the cost of the enclosed region |
| US5903664A (en) * | 1996-11-01 | 1999-05-11 | General Electric Company | Fast segmentation of cardiac images |
| WO2003021532A2 (en) * | 2001-09-06 | 2003-03-13 | Koninklijke Philips Electronics N.V. | Method and apparatus for segmentation of an object |
| US20040008886A1 (en) | 2002-07-02 | 2004-01-15 | Yuri Boykov | Using graph cuts for editing photographs |
| US7403211B2 (en) * | 2003-02-13 | 2008-07-22 | Lumapix, Inc. | Method and system for interactive region segmentation |
| WO2004077358A1 (en) * | 2003-02-28 | 2004-09-10 | Cedara Software Corporation | Image region segmentation system and method |
| DE102004027710A1 (en) | 2004-06-07 | 2006-01-26 | Siemens Ag | Method for the automatic detection of a structure in medical imaging, computer tomography device, workstation and computer program product |
| US7339585B2 (en) * | 2004-07-19 | 2008-03-04 | Pie Medical Imaging B.V. | Method and apparatus for visualization of biological structures with use of 3D position information from segmentation results |
| WO2006056614A1 (en) * | 2004-11-27 | 2006-06-01 | Bracco Imaging S.P.A. | 2d / 3d integrated contour editor |
| US7676081B2 (en) * | 2005-06-17 | 2010-03-09 | Microsoft Corporation | Image segmentation of foreground from background layers |
| US20070165966A1 (en) * | 2005-07-15 | 2007-07-19 | Yissum Research Development Co. | Closed form method and system for matting a foreground object in an image having a background |
| US7822274B2 (en) * | 2006-01-17 | 2010-10-26 | Siemens Medical Solutions Usa, Inc. | Banded graph cut segmentation algorithms with laplacian pyramids |
| US8233712B2 (en) * | 2006-07-28 | 2012-07-31 | University Of New Brunswick | Methods of segmenting a digital image |
| US20080298544A1 (en) * | 2007-05-29 | 2008-12-04 | Peter Dugan | Genetic tuning of coefficients in a threat detection system |
| US8009911B2 (en) * | 2007-08-30 | 2011-08-30 | Siemens Aktiengesellschaft | Interactive image segmentation on directed graphs |
-
2007
- 2007-06-22 DE DE102007028895A patent/DE102007028895B4/en not_active Expired - Fee Related
-
2008
- 2008-06-18 CN CN2008101253079A patent/CN101329763B/en not_active Expired - Fee Related
- 2008-06-18 US US12/213,322 patent/US8180151B2/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1524247A (en) * | 2001-05-17 | 2004-08-25 | О | A variational approach for the segmentation of the left ventricle in mr cardiac images |
| CN1748641A (en) * | 2004-08-09 | 2006-03-22 | 西门子公司 | Method for segmenting a medical data set |
| CN1745715A (en) * | 2004-09-09 | 2006-03-15 | 西门子公司 | Method for segmenting anatomical structures from 4d image data |
Also Published As
| Publication number | Publication date |
|---|---|
| US8180151B2 (en) | 2012-05-15 |
| DE102007028895B4 (en) | 2010-07-15 |
| DE102007028895A1 (en) | 2009-01-02 |
| CN101329763A (en) | 2008-12-24 |
| US20080317342A1 (en) | 2008-12-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101329763B (en) | Method for segmenting structures in image data records and image processing unit | |
| US8160320B2 (en) | Medical image display apparatus, method and program, and recording medium for the program | |
| EP3035287B1 (en) | Image processing apparatus, and image processing method | |
| US10885392B2 (en) | Learning annotation of objects in image | |
| EP2024937B1 (en) | Method and system for generating a 3d representation of a dynamically changing 3d scene | |
| US9710880B2 (en) | User-guided shape morphing in bone segmentation for medical imaging | |
| CN102246208B (en) | Image segmentation | |
| US9436798B2 (en) | Method of retrieving data from a medical image data set | |
| CN103402435B (en) | Medical image processing device and medical image processing method | |
| US20080109717A1 (en) | Reviewing editing operations | |
| US7386153B2 (en) | Medical image segmentation apparatus and method thereof | |
| US20110234630A1 (en) | Active Overlay System and Method for Accessing and Manipulating Imaging Displays | |
| US20060241412A1 (en) | Method for visualizing damage in the myocardium | |
| US20100150418A1 (en) | Image processing method, image processing apparatus, and image processing program | |
| US20160086049A1 (en) | Contour correction device, method, and program | |
| US10013810B2 (en) | Contouring tool having automatic interpolation and extrapolation | |
| CN101479762A (en) | A method, a system and a computer program for determining a threshold in an image comprising image values | |
| US20150317799A1 (en) | Medical image processing apparatus, method, and program | |
| US20170027536A1 (en) | Apparatus and method for reconstructing panoramic x-ray image | |
| EP2057573B1 (en) | Presentation method, presentation device and computer program for presenting an image of an object | |
| US9530238B2 (en) | Image processing apparatus, method and program utilizing an opacity curve for endoscopic images | |
| US8306354B2 (en) | Image processing apparatus, method, and program | |
| JP2009080660A (en) | Object region extraction processing program, object region extraction device, and object region extraction method | |
| US8200015B2 (en) | Method for interactively segmenting structures in image data records and image processing unit for carrying out the method | |
| EP2761590A1 (en) | Method and system of defining a region of interest on medical scan images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130102 Termination date: 20150618 |
|
| EXPY | Termination of patent right or utility model |